1. A Clinical Trial Disaster in UK
February, the 16th 2007
A Clinical Trial Disaster in UK
Laura BATIQUE,
Kelly BROWN,
Céline DELPLACE,
Bénédicte PALADINI,
& Pauline SALADIN

2. The « TeGenero matter » in few words
Introduction
The « TeGenero matter » in few words

3. Description Clinical trial : phase 1
TGN 1412 : CD28 superagonist monoclonal antibody
Role
Name
Location
Sponsor
TeGenero
Germany
Manufacturer
Boehringer Ingelheim
Contract Research Organisation
Parexel
Northwick Park Hospital, UK

4. What's happened ? Phase 1 : Cytokine storm, lymphopenia, monocytopenia
Six healthy volunteers received TGN 1412 at Northwick Park Hospital
Phase 1 : Cytokine storm, lymphopenia, monocytopenia
Phase 2 : Reactive phase, multiorgan failure
Phase 3 : Recovery phase
Phase 4 : Steady state
13th March 2006

5. PLAN Part I : Media’s reaction
Part II : Scientific context & mechanism of action
Part III : Preclinical studies
Part IV : Transition to clinical development
Part V : Clinical trial
Part VI : Protagonist reaction
Part VII : Discussion

6. Part I : Media’s Reaction

7. MEDIA
Telegraph.co.uk

8. Media’s reaction Some days after the trial
Violent Reaction to monoclonal antibody therapy remains a Mystery (Science 15/03/06)
Tragic drug trial spotlights potent molecule (Nature 13/03/06 )
« There are a number questions that need answering […] what happened […] monkeys were an appropriate animal model to test […] TeGenero had tested the antibody on human blood […] why Parexel did not leave a longer gap between doses » (Pharmatimes 26/05/06)
British trial disaster casts doubts on testing guidelines (Nature 01/04/06 )

9. What caused this event ?
Errors in manufacture, formulation or administration?
Contamination with endotoxin, pyrogen, microbiologic or other agents?
Mechanism of action?

10. Part II : Scientific context & mechanism of action

11. Costimulation of Lymphocyte T

12. The costimulation
T-Cell Costimulation — Biology, Therapeutic Potential, and Challenges Arlene H. Sharpe, M.D., Ph.D., and Abul K. Abbas, M.D.

13. Interaction with costimulary signals
induce tolerance
to treat auto-immune diseases and prevent graft rejection
booste immune response
to enhance antimicrobial response and antitumoral response
Improve the action of CD28
Inhibit the action of CTLA4
Inhibit the action of CD28
Improve the action of CTLA4

14. Abatacept Orencia® Ipilimumab and ticilimumab
The first target: CTLA4
Abatacept Orencia®
Ipilimumab and ticilimumab

15. In this context of « sucessful » development of drugs targeting costimulary molecules, a new mecanism has been found.

16. The second target: CD28 What is a superagonist antibody?
The first indication: leukemia
The second indication: autoimmune diseases
Risks of TGN1412

17. TGN 1412 CD 28 superagonist antibody
Humanized
Ig G 4
Superagonist : no costimulation
TCR
CD28
T cell

18. The first indication: chronic lymphocytic leukemia

19. What is chronic lymphocytic leukemia?
Accumulation of leukemic B cells
T cell abnormalities
Increase the risk of infection
Tolerance versus leukemic B cells: hamper the recognition and elimination of leukemic B cells

20. How does TGN1412 work?
T-cell
Activation of T cells stoppes the tolerance to the leukemic B cells
CD3
T-cell Signals
TCR
CD4/CD8
Costimulatory Signals
CD28
CD28 Superagonist
Elimination of leukemic B cells

21. The second indication: autoimmune diseases

22. Auto-immune diseases Generation of self-reactive T cells
Regulatory T cells are not efficients

23. How does a CD28 superagonist work?
What are regulatory T cells?
Theoric effect of TGN1412 on the Treg

24. Costimulatory Molecules
Production of regulatory T cells (Treg)
In the Thymus
T-cell
Thymic cell
Immunologic Synapse
MHC
TCR
CD4/CD8
LFA-3
CD2
CD40
CD40L B7
CD28
Costimulatory Molecules
CD3
Costimulatory Signals
Self antigen
T-cell Signals
Regulatory T cells are autoreactives

25. Regulatory T cells are autoreactives
Autoimmune process
Lesion of tissue
Self-antigen are released
APCs collect the self-antigen
presentation of the self-antigen to the Treg
Activation of Treg
Suppressive activity on the others T cells

26. Functional capacities of Tregs:
a suppression mediated by cell-to-cell contact
Autoimmunity: Basic Mechanisms and Implications in Endocrine Diseases Part II S. Ballotti F. Chiarelli M. de Martino Department of Paediatrics, Anna Meyer Children’s Hospital, University of Florence, Florence , and Department of Paediatrics, University of Chieti, Chieti , Italy

27. Effect of TGN1412 Activation of regulatory T cells (Treg)
CD3
T-cell Signals
TCR
CD4
CD25
Costimulatory Signals
CD28
CD28 Superagonist
activation of Treg

28. Risks of a CD28 superagonist

29. Presence of CD28 molecules
CD 28 is present at surface of CD4+ T cells: Treg and others lymphocytes T4
CD28 is present on the half of lymphocytes CD8+ and granulocytes
Selectivity of TGN 1412 ????????
Controlled activation of the cells
Establishing risk of human experimentation with drugs: lessons from TGN 1412; M J H Kenter, A F Cohen; the lancet; october 2006

30. Superagonist Antibodies ?

31. C’’D loop X-ray cristallographic analyses
Agonistic anti-CD28 monoclonal abs:
 bind exclusively to a specific part of the CD 28 molecule: the laterally exposed C’’D loop of the Ig-like extracellular domain of CD 28
Specificity closely correlated with the superagonistic activity of anti-CD28 antibodies
CD 28 with 5.11A1

32. X ray cristallographic structure
(3D model: extracellular part of human CD28)
MYPPPY motif (aa ) for B7 binding (natural ligand)
Adjacent aa 98 residue for binding of conventional CD28-specific mAb
C’’D loop (aa 60-65) for the binding of superagonistic CD28-specific mAb
Topological Requirements and Signaling Properties of T Cell–activating, Anti-CD28 Antibody Superagonists
Fred Lühder,Yun Huang,Kevin M

33. Part III : Preclinical Studies (Toxicology)
1.Rodent: rat
2.Non human primate: monkey

34. 1.Toxicology On Rats

35. Preclinical On Rats
Presentation of molecules
Trials on healthy rats in vivo
Trials on unwell rats in vivo

36. Presentation of molecules
Domaine antiCD28 humain obtenu à partir du rat ←
CD28 rat defers by 68% with CD28 human :
Need of a Ab anti-CD28
specific for the rat : JJ316
Monoclonal antibody anti-humanCD28
Domaine constante humain ←

37. Trials on healthy rats The 2 phenotypes of CD4+Tcells
Identification of regulatory Tcells (Treg)
JJ316’s action on CD4+Tcells

38. 2 phenotypes are possible for CD4+Tcells : CD25+ ou CD25-
The phenotypes
2 phenotypes are possible for CD4+Tcells : CD25+ ou CD25-
Experiment :
- isolation of Tcells lymph nodes from Lewis rats
- marking : CD4+ with CFSE
CD25+ with anti-CD25 mAb

39. Observation : flow cytometric
CD4
CD25

40. Identification of Treg
Experiment :
- Western Blot and flow cytometric
- detection of FOXP3 on the Treg
Observation : FOXP3 is only expressed on CD25+
Conclusion : Treg’s phenotype is : CD4+CD25+

41. ↑ of Treg, but maybe other cells
FoxP3 is an intracells marker, specific of Tregcells
BUT, we need an extracells marker to appreciate the rate of cells ! → CD25+
Pb : CD25+ not specific of Treg
↑ of Treg, but maybe other cells
CD25+
CD25-

42. Ab’s action Aim : to quantify the JJ316’s action Experiment :
- injections of several JJ316 or MOPC-31C dosages to Lewis rats (0.1 à 1mg/kg)
- isolation of lymph node and spleen
- flow cytometric

43. JJ316 ↑ both cells

44. Conclusion :
- 0.1 mg : significant ↑ of Treg
- 1 mg : significant ↑ of Treg and Tconv
a low-dose of the JJ316 is sufficient for increasing Treg without increasing Tconv

45. Abstract
Identification of different CD4+ Tcells (due to their phenotype) : Treg and Tconv
To show clearly the selective ↑ of Treg with dosage ≤ 0.3 mg/kg of JJ316
So don’t make mistakes in doses !

46. Trials on unwell rats Experiment : determination of the DME
- treatment of rats with different doses : 0.03 / 0.1 / 0.3 mg/kg of JJ316
- evaluation of clinical symptoms

47. The weakest
clinical
symptoms is
obtained with
0.3 mg/kg

48. Abstract Preferential ↑ of Treg at low-dose (0.1-0.3)
Moreover : Treg ↑ IL10
Teff ↓ IL2 and IL4
No Problem concerning the rats

49. 2.Toxicology On Monkeys and Limits
In vitro & Failure in Investigator’s brochure
In vivo
Pharmacokinetics & Toxicokinetics

50. In vitro: animal and human cells
Total Peripheral Blood Mononuclear Cells (PBMC)
Anti-human CD28 agonistics:
- 5.11A1 = parental antibody (TGN 1412: humanised
version of the mouse antibody 5.11A1)
- TGN 1112 (IgG1 variant)
- TGN 1412 ( IgG4)
Share specificity-function relations: C’’D loop

51. Selection of appropriate species  Safety and toxicology studies
Non-Human primates:
Rhesus monkey
Cynomolgus monkey
Marmoset monkey
Selection of appropriate species
 Safety and toxicology studies
TGN 1412 reactive with human, cynomolgus, rhesus T cells

52. Failure in Investigator’s brochure
No comparison result: activation of human/monkey PBMC
However, stimulation of Cynomolgus macaques and human PBMC with TGN 1412:
Dose of TGN 1412 given to volunteers = close to the maximum immunostimulatory dose

53. Comparison of the human and rhesus monkey CD 28 aminoacid sequences:
Investigator: 100% identity between human and monkey CD28 sequences
But no sequence comparison !
In fact some differences:
Comparison of the human and rhesus monkey CD 28 aminoacid sequences: *
Establishing risk of human experimentation with drugs: lessons from TGN 1412; M J H Kenter, A F Cohen; the lancet; october 2006

54. Amino acid sequences homology
C’’D loop:
Non conservative variation at position 65 in rhesus monkey
TGN 1412  different bindings
 different T cells activation
Amino acid sequences homology
(C’’D loop)
Human
Cynomolgus
identical
Rhesus
Differs in 1 aa
Marmoset
30%

55. Failure in Investigator’s brochure
No test done to demonstrate identity CD28 sequences between cynomolgus monkeys and humans
No comparison between rhesus, cynomolgus monkeys and human CD28 sequences
No comparison: affinity of TGN 1412 on human and monkey CD28

56. In vivo: animal models Rhesus monkey:
- TGN 1112  Activation and expansion of Tcells
- TGN 1412  significantly weaker pharmaceutical activity
Cynomolgus monkey:
- TGN 1412  Activation and expansion of T cells
- Studies conducted in this monkey!
No comparison result between the action of TGN 1412/TGN 1112 !!

57. Pharmacokinetics and Toxicokinetics
Cynomolgus monkey: the most predictive
Repeat-dose toxicity study: - IV
- 5 up to 50mg/kg
TGN 1412 : well tolerated up to 50mg/kg/week
 50mg/kg = NOAEL (No-Observed-Adverse-Effect Level)
Half life = 8 days
 Full removal = 1 month

58. Part IV : Transition to Clinical Development of TGN 1412
Differences between mouse/rat models & humans
Affinity differences between antibodies
Use of healthy volunteers
Dose calculation

59. Differences between mouse/rat models and humans
Preclinical studies (mice) :
Work in pathogen-free conditions
 majority of naive T cells
Humans :
Microbe-rich natural habitat
 majority of memory T cells
 Activation of memory easier than naive T cells
 TGN 1412 : In humans, widespread immune activation not in mice

60. Affinity Differences between Antibodies
Differents Abs: JJ 316 # TGN 1112 # TGN 1412 # 5.11A1
 Pharmacological activity in rhesus monkey:
TGN 1412 << TGN 1112
Anti-Human-CD28 Ab:
- weak agonist in monkeys
- Strong agonist in humans
 Stimulation of T cells: in humans >> in primates
Target: (sub)epitopes probably present only in humans

61. Why did the compagny choose healthy volunteers?
No adverse effects in healthy animals
No pre-activation or dysfunction of T cells
No pre-existing imbalance: Tconv/Treg cells
Homogenous population:
 No impact of pre/co-medication and/or disease activity
No interference with interpretation of TGN 1412 safety and pharmacology

62. Choice of patients would be better!
Case by case
Pharmacokinetic (mAb):
antigen dose
pathological state
Target epitopes
Mechanism of action
 TGN 1412: Highly specific
Responses of immune system: healthy volunteers # patients
A lack of information!!
Treatment targets immune system

63. How was the starting dose calculated?
Starting dose = 0,1mg/kg
NOAEL = 50mg/kg
10 (safety)
+ Safety margin
3,1
HED = 16mg/kg
MSRD = 1,6mg/kg
Human Equivalent Dose
Maximum Recommended Starting Dose
 Normally ensure maximum patient safety

64. What precautions should have been done ?
« MABEL » Minimal Anticipated Biological Effect Level
mg/kg
2,5 25 50
Cynomolgus Rhesus monkeys ?
pharmacological activity
NOAEL
mg/kg
0,3 1 5
MABEL
healthy & arthritic rats
NOEL
optimal pharmacological responses
+Safety criteria
(CPMP)
Safe starting dose
= 0,005mg/kg!!
If MABEL = 0,5mg/kg

65. Calculated receptor occupancy with TGN 1412
Example of receptor occupancy versus local concentration:
Dose (mg/kg)
Receptor occupancy (%)
0,0001
0,6
0,001
5,9
0,01
38,4
0,1
86,2 1
98,4 10
99,8
Initial dose:
very high occupancy of CD28 receptors
likely to achieve maximum pharmacological effect

66. Part V : Clinical Development
Trial protocol
Adverse effects observed
Consequences
Failure in the protocol

67. Trial Protocol

68. Design of trial Single-centre Double-blind Randomized
Placebo-controlled
4 groups of 8 healthy young male volunteers planned
Administration : single doses intravenously
Dosing range : 0.1, 0.5, 2.0 and 5.0 mg/kg BW
Dosing time : 8 to 10 a.m
First day : 13th March 2006

69. Volunteers No notable medical history Informed consent
£2000 fees + £30 per visit
First group of eight volunteers :
19 – 34 years old
Six received active agent
Two received placebo

70. Objectives of trial Primary : Assessment of the safety & tolerability
Determination of the pharmacokinetics
Secondary :
Determination of the effect on lymphocyte subsets
Assessment of the cytokine profile
Assessment of anti-TGN1412 antibodies

71. Approval
27th January 2006 : Authorization by the Medicines and Healthcare products Regulatory Agency (MHRA)
14th February 2006 : Favourable opinion by the Brent Medical Ethics Committee

72. Adverse Events observed

73. Pulmonary infiltrates Disseminated intravascular coagulation
Main adverse events
13th March 2006
8 to 9 a.m
TGN infusion
0.1 mg/kg
Desquamation, peripheral ischemia
Multiorgan failure 1h
1h30
12 to 16h
24h
2 days
10 days
30 days
Admittance to the intensive care unit
Severe headache
Several months after the event : inability to concentrate, headaches, loss of fingers & toes.
Cytokine release syndrome
Lymphopenia
Monocytopenia
Respiratory distress
Pulmonary infiltrates
Renal failure
Disseminated intravascular coagulation

74. Cytokine release syndrome
TNF α
IFN γ
IL-2
Source: EXPERT SCIENTIFIC GROUP ON PHASE ONE CLINICAL TRIALS
(20th July 2006)

75. Common features Striking phenomenon = Stereotypical response
in all six patients
in all organ systems affected
Source : (September 7,2006)

76. Lymphopenia & monocytopenia

77. Consequences

78. Treatment Initial : vasoconstrictor, analgesics, hydrocortisone
In Critical Care :
Ventilation
Hemodiafiltration
Hydrocortisone followed by methylprednisolone IV
Daclizumab
Treatments for infections/viral reactivation
Freshfrozen plasma & cryoprecipitate
Irradiated red cells & platelets

79. Measures taken by MHRA The MHRA :
Incidents reported to the MHRA on the afternoon of 14th March
The MHRA :
Immediately suspended the authorization
Confirmed that the drug was not in use in any other trial
Alerted international drug regulatory authorities
Sent a team of inspectors to the unit in Northwick Park
MHRA = The Medicines and Healthcare products Regulatory Agency

80. Failure in the Protocol

81. Doses administered and timing
Subjects 1 2 3 4 5 6
Body weight (kg)
84.3
68.9
88.5
82.4
72.1
81.8
Dosing TGN 1412
8.4mg
6.8mg
8.8mg
8.2mg
7.2mg
Dosing time
08:00
08:20
08:30
08:40
08:50
09:00
Source: EXPERT SCIENTIFIC GROUP ON PHASE ONE CLINICAL TRIALS (20th July 2006)

82. Comments Authorization by MHRA
All subjects are dosed at the same session
 Staff insufficient
Two placebo
 Two sessions ?
Subjects were dosed at ~ 10 minute intervals
 ↑ intervals ?
BUT longer therefore more expensive ?

83. Part VI : Protagonist Reaction
Boehringer Ingelheim
Parexel
TeGenero

84. Boehringer Ingelheim Sorry for the patients of trial
Contract manufacturer, the material for investigational use only
Conformed to GMP
No contamination

85. Parexel Sorry for the patients of clinical trial
To assist the pharmaceutical, biotech, and medical device industries Investigator
No deficiency during inspection by MHRA
Good practices and policies and procedures
MHRA confirmed the approved protocole

86. Te Genero’s Team Sorry for trial patients, their family
« Develop innovative treatment to help millions of people »
« Side effects could have been caused by one of our products »
MHRA agreement for protocole used
« No sign of risk from preclinical test »

87. The victims Tom Edward: lucky escape from trial
Myfanwy Marshall’s boyfriend
Msg de la diapo: La majorité des patients étaient jeunes et avaient entre 20 et 30 ans.
Il y avait des étudiants qui participaient pour se payer leur étude ou le permis pour le boulo comme Ryan Wilson
Meme les patients témoins ont eu tres peur tom edward raconte qu’il ne savait ce qu’il allait lui arriver quand il a vu peu à peu les 6 jeunes hommes déclarer des effets secondairess.
Les familles furent aussi touchées car quand elles ont été prévenu ils étaient tous en soin intensifs et leurs jours étaient encore comptés.
Ryan Flanagan: 21, a student from Highbury
Ryan Wilson: 20, the most seriously ill

88. The consequences Ryan Wilson: 20, the most seriously ill
Heads and bodies swelled up
He was suffering from heart, liver and kidney failure, pneumonia and septicaemia.
Development gangrene Amputation parts on his fingers
Risk infections and cancer
Rapidement extrémités froides des mains et douleurs
Il ne pourra jamais atteindre son ambition qui été de devenir plombier.

89. Process indemnity in progress Rejected by victim’s lawyers
Indemnities
Phase I
Accident
13 March
April 2006
2007
2 900€
Before safety trial
7 500€
Each of the victims (Tegenero)
15 000€
No condition
Process indemnity in progress
For accepting a no fault agreement
Rejected by victim’s lawyers

90. Part VII : Discussion Solutions to find

91. ESG’S Objectives (20/12/06) To put in position new
recommendations to OPTIMISE
the SECURITY of Clinical Trials’Phase I

92. Recommendations
Characterisation of new agents’risks and Harmonisation between all authorities
Choice of preclinical and clinical models
Doses’schema

93. Wider approach for doses : the MABEL
Staff & Environment
Better communication for intra and inter- tests

94. Creation of the EAG
Time-Scale
Creation of Specialised Centres
Creation of Specialist Centres

95. Conclusion

96. TeGenero’s bankcruptcy
Catastrophic safety trial
To file for Involvency
Bankcruptcy
March 2006
April 2006
4 July 2006
December 2006
13 March
Tegenero
Need result
Victim’s indemnisation
Need to attract the investment
Anonymous investor wants to buy TeGenero
Impossible
Negative attention €
Insurance
Obtain money by Investors

98. Hardening of the system
CONCLUSION
Could this disaster have been avoided?
Yes
The culpables & sanctions
In process
The lessons to remember:
Need changing and
Hardening of the system
More controls
Les recommandations ne suffisent pas

99. Merci de votre attention

100. Powerpoint bonus

101. soluble fraction of CTLA4 + Ig G Fc region
abatacept Orencia®
soluble fraction of CTLA4 + Ig G Fc region
Fixation on B7 on peripheral antigen-presenting cells
HLA
TCR
CD28 can’t interact with B7 anymore B7
abatacept
CD28
inhibition of the activation of the T cells
Peripheral antigen-presenting cell
T cell
Approved in 2005 for the treatment of rheumatoid arthritis and in development for other indications

102. Adverse events:
Phase 4 trial:
One year, randomized, multicenter, double-blind, placebo-controlled trial
Serious infections were more frequent in the abatacept group than in the placebo group (2.9% versus 1.9%).
 Safety of the selective costimulation modulator abatacept in rheumatoid arthritis patients receiving background biologic and nonbiologic disease-modifying antirheumatic drugs: A one-year randomized, placebo-controlled studyM. Weinblatt 1 *, B. Combe 2, A. Covucci 3, R. Aranda 3, J. C. Becker 3, E. Keystone 4

103. Antibody anti-CTLA4: ipilimumab, ticilimumab
Blocking antibody anti-CTLA4: No interaction between B7-CTLA4
TCR
HLA
Promotes interaction between B7-CD28 (and HLA-TCR)
CTLA4 B7
CD28
Lenghtens the T cells activation
antigen-presenting cell APC
T cell
Currently in development for the treatment of tumoral disease

104. Efficacy of a human anti-CTLA4 antibody + IL2: a phase 1/2 study
Patients: metastatic melanoma
Treatments:
36 patients received injection of ipilumimab at differents doses every 3 weeks
All patients received IL2 therapy
Efficacy:
8 patients (22%) experiences positive response
6 of the 8 have ongoing the response at 11 to 19 months
Tumor regression and autoimmunity in patients treated with cytotoxic T lymphocyte-associated antigen 4 blockade and interleukin 2: a phase I/II study. Ann Surg Oncol Dec;12(12): Epub 2005 Oct 21

105. Adverse effects
5 patients (14%) developed autoimmune toxicities secondary to anti-CTLA4 administration:
- 4 with enterocolitis
- 1 with arthritis and uveitis
Edema and ulcerations in the descending colon of a patient
Tumor regression and autoimmunity in patients treated with cytotoxic T lymphocyte-associated antigen 4 blockade and interleukin 2: a phase I/II study. Ann Surg Oncol Dec;12(12): Epub 2005 Oct 21